Visual acuity outcomes in pediatric glaucomasMonica R. Khitri, MD, Monte D. Mills, MD, Gui-Shuang Ying, PhD, Stefanie L. Davidson, MD,and Graham E. Quinn, MD

PURPOSE To evaluate and compare the visual acuity prognosis in the various pediatric glaucoma

Author affiliations: DepartmentScheie Eye Institute, PhiladelphiSubmitted February 16, 2012Revision accepted May 1, 201Correspondence: Monica R. K

Woodland Ave., Philadelphia, PCopyright � 2012 by the Am

Strabismus.1091-8531/$36.00http://dx.doi.org/10.1016/j.ja

376

subtypes and to determine risk factors for vision loss.

METHODS The medical records of pediatric glaucoma patients from 2000 to 2010 at Children’s

Hospital of Philadelphia were retrospectively reviewed. Visual acuities, surgeries, glau-coma subtype, and etiology of vision impairment were recorded. Univariate and multivar-iate analyses were performed to determine the risk factors for visual impairment.

RESULTS A total of 133 eyes (36.8% primary congenital glaucoma, 28.6% aphakic glaucoma, 12.0%

glaucoma associated with anterior segment dysgenesis, 12.0% Sturge-Weber glaucoma) of88 patients were included. At last follow-up (median length, 5 years), 46.6% eyes achievedexcellent ($20/70) visual acuity. Of the glaucoma subtypes, primary congenital glaucomaconferred the best visual prognosis, with 69.4% eyes with excellent ($20/70) visual acuityat final follow-up. Factors most associated with visual impairment (\20/200) were unilat-eral disease, multiple surgeries, poor vision at diagnosis, and other ocular comorbidities.The most common primary etiology for vision impairment was amblyopia (54.9%).

CONCLUSIONS Patients with glaucoma early in life appear to have a better visual acuity prognosis than pre-

viously reported, with those with primary congenital glaucoma faring better than otherglaucoma subtypes. Recognition of risk factors for visual impairment can better guideclinical management and counseling of patients. ( J AAPOS 2012;16:376-381)

Pediatric glaucoma is a challenging condition totreat and is commonly associated with extensive vi-sion impairment. The etiology of patients’ vision

impairment is often multifactorial, including amblyopia,optic nerve damage, and other ocular comorbidities.1-3

Recent classification schemes of childhood glaucomashave focused on distinguishing the various pediatricglaucomas by etiology, thereby assisting in the selectionof treatment choices specific to the given condition.4 Forinstance, primary congenital glaucoma often is treatedwith anterior segment angle surgery such as goniotomy ortrabeculotomy as the first-line management strategy,whereas secondary glaucomas, such as those associatedwith anterior segment dysgenesis and aphakic glaucoma,may be initiallymanagedmedically. A better understandingof how the visual prognosis compares between the variouspediatric glaucoma subtypes would aid not only in counsel-ing parents at the time of diagnosis but also in tailoringtreatment plans to address better an individual patient’sgreatest risk factor for vision impairment. The purpose of

of Ophthalmology, Children’s Hospital of Philadelphia,a, Pennsylvania.2.hitri, MD, Philadelphia VA Medical Center, 3900A 19104 (email: monica.ralli@gmail.com).erican Association for Pediatric Ophthalmology and

apos.2012.05.007

this study was to determine the relationship among the var-ious subtypes of pediatric glaucoma and their visual out-comes and to identify the factors that most affected visualprognosis.

Methods

The medical records of all patients with pediatric glaucoma

treated between 2000 and 2010 at Children’s Hospital of Phila-

delphia were retrospectively reviewed. Institutional review board

approval was obtained for data review and analysis, and the study

conformed to the requirements of the Health Insurance Portabil-

ity and Accountability Act. Eligible patients were identified

through queries of the medical billing record. Exclusion criteria

for the study included age .14 years at time of diagnosis, glau-

coma related to trauma, and diagnosis and/or initial management

of glaucoma made at another hospital.

The following data were collected: type and laterality of glau-

coma, age at diagnosis, ages at all surgeries, types of surgeries un-

dertaken, visual acuities, and associated ocular and systemic

comorbidities. Visual acuities were measured using fixation or

Teller grating acuity cards for nonverbal children and Lea sym-

bols, HOTV, and Snellen letters progressively for verbal children

as cooperation allowed. For those patients with subnormal final

visual acuities (\20/70) the primary etiology for the poor vision

also was determined. The cause of vision impairment in glaucom-

atous eyes can be multifactorial; however, we determined the sin-

gle most important contributing factor for each eye’s loss of vision

on the basis of the physical examination (ie, presence of optic

nerve pallor and/or cupping, retinal degeneration) and history

Journal of AAPOS

Table 1. Characteristics of study subjects with pediatric glaucoma(N 5 133 eyes)

Age at diagnosis (yrs)Mean (SD) 2.11 (3.35)Median (min, max) 0.36 (0.0, 13.8)

Type of glaucoma, n (%)Primary congenital glaucomaa 49 (36.8)Aphakic glaucomab 38 (28.6)Glaucoma associated with anterior segmentdysgenesisc

16 (12.0)

Sturge-Weber glaucoma 16 (12.0)Uveitic glaucomad 5 (3.76)Aniridia associated glaucomae 4 (3.01)Otherf 5 (3.76)

Ocular comorbidities at diagnosis, n (%)Cloudy cornea 53 (39.9)Aphakia 38 (28.6)Anterior segment dysgenesis 16 (12.0)Enlarged cornea 5 (3.76)Uveitis 5 (3.76)Aniridia 4 (3.01)Nystagmus 4 (3.01)Strabismus 2 (1.50)Retinal degeneration/injury 1 (0.75)ROP 1 (0.75)Cataract 1 (0.75)Pseudophakia 1 (0.75)

Systemic comorbidities at diagnosis, n (%)Developmental delay 9 (10.2)Chromosomal abnormality 5 (5.7)Prematurity 5 (5.7)Juvenile idiopathic arthritis 4 (4.6)Sickle cell disease 2 (2.3)

Vision at diagnosis, n (%)Excellent (20/20 to 20/70, CSM) 35 (26.3)Good (\20/70 to 20/200, CUSM) 73 (54.9)Fair (\20/200 to 20/800, CUSUM, CSUM) 3 (2.26)Poor (\20/800 to light perception, UCSUM,UCUSUM)

20 (15.0)

No light perception 2 (1.50)Age at last follow-up, yearsMean (SD) 9.24 (6.87)Median (min, max) 7.05 (1.28, 34.3)

Years of follow-up since diagnosisMean (SD) 7.14 (6.18)Median (min, max) 4.95 (0.29, 27.9)

CSM, central-steady-maintained; CSUM, central-steady-unmaintained;CUSM, central-unsteady-maintained; CUSUM, central-unsteady-unmaintained; ROP, retinopathy of prematurity; UCSUM, uncentral-steady-unmaintained; UCUSUM, uncentral-unsteady-unmaintained.aNewborn primary congenital glaucoma, infantile primary congenitalglaucoma, or late-recognized primary congenital glaucoma.4bInfantile aphakic open-angle glaucoma.4cSclerocornea, Peter’s anomaly, iridotrabecular dysgenesis (irishypoplasia) syndrome, or Axenfeld-Reiger syndrome.4dOpen-angle glaucoma associated with chronic uveitis.eCongenital aniridic glaucoma or acquired aniridic glaucoma.fAngle-closure glaucoma secondary to retinopathy of prematurity,glaucoma associated with congenital ocular melanosis, or juvenileopen-angle glaucoma.

Table 2. Management in pediatric glaucoma (N 5 133 eyes)

Surgical management, n (%)No 9 (6.77)Yes 124 (93.2)

Age at initial surgery, yra

Mean (SD) 2.41 (4.19)Median (min, max) 0.55 (0.01, 19.8)

Age at most recent surgery, yra

Mean (SD) 4.03 (4.80)Median (min, max) 1.62 (0.01, 28.3)

Initial surgery performed, n (%)a

Angle surgery 89 (66.9)Filtering surgery 20 (15.0)Cyclophotocoagulation 11 (8.30)Peripheral iridectomy 2 (1.5)Other 2 (1.5)

Most recent surgery performed, n (%)a

Angle surgery 44 (33.1)Filtering surgery 38 (28.6)Cyclophotocoagulation 21 (15.8)Revision of filtering surgery for failure 8 (6.0)Revision of filtering surgery for malpositionedtube

6 (4.5)

Other 7 (5.3)Total number of surgeries per eyea

1 46 (37.1%)2 37 (29.8%)3 14 (11.3%)4 13 (10.5%)5 3 (2.4%)6 4 (3.2%)7 3 (2.4%)8 2 (1.6%)10-12 2 (1.6%)Mean (SD) 2.35 (2.01)Median 2

aRestricted to those that received surgery management.

Volume 16 Number 4 / August 2012 Khitri et al 377

(compliance with patching regimens, changes in visual acuity over

time). IOP measurements were not investigated in this study,

given the variability of these measurements with patient coopera-

tion in this age group.

Journal of AAPOS

Patientswere clinicallymanaged by 1 of 3 pediatric ophthalmol-

ogists (MDM, SLD, GEQ). The surgeries were classified as angle

surgery (goniotomy, trabeculotomy), filtering surgery (trabecu-

lectomy, placement of an aqueous drainage device), revision of fil-

tering bleb, cyclophotocoagulation, peripheral iridectomy or

iridotomy, or other (anterior chamber washout, synechiolysis, re-

vision of tube shunt for malposition). The aqueous drainage de-

vices used were the Ahmed glaucoma valve implant (New World

Medical Inc, Rancho Cucamonga, CA), the Baerveldt implant

(Advanced Medical Optics, Santa Ana, CA), and the Molteno im-

plant (Ophthalmic Ltd, Dunedin, New Zealand). The Ahmed

valve implant was placed in a single procedure. The Baerveldt

and the Molteno implants were placed in a staged process with

a 4- to 6-week interval between the plate placement and the tube

introduction. For the purposes of this study, these two-staged pro-

cedures were counted as one surgery; the date recorded was that of

the second procedure given that the IOP-lowering effect of the

surgery would not occur until after the second procedure. Besides

surgery, patients often were on topical or systemic glaucomamed-

ications during the course of their treatment.

Statistical Analysis

The patients’ ocular characteristics at diagnosis and characteris-

tics of surgical management were summarized by mean, SD,

Table 3. Final visual outcomes in pediatric glaucoma (N 5 133eyes)

Visual outcomes n (%)

Visual acuity at last follow-upExcellent (20/20 to 20/70, CSM) 62 (46.6)Good (\20/70 to 20/200, CUSM) 18 (13.5)Fair (\20/200 to 20/800, CUSUM, CSUM) 12 (9.02)Poor (\20/800 to LP, UCUSUM, UCSUM) 35 (26.3)NLP 6 (4.51)

Functional visual outcome at last follow-upa

Good vision in both eyes 23 (35.9)Good vision in one eye 31 (48.5)Good vision in neither eye 10 (15.6)

Primary etiology of vision lossb

Amblyopia 39 (54.9)Glaucoma 18 (25.4)Retinal disease 10 (14.1)Corneal opacification 2 (2.82)Cataract 1 (1.41)Other optic neuropathy 1 (1.41)

CSM, central-steady-maintained; CSUM, central-steady-unmaintained;CUSM, central-unsteady-maintained; CUSUM, central-unsteady-unmaintained; UCSUM, uncentral-steady-unmaintained; UCUSUM,uncentral-unsteady-unmaintained.aAmong those who had data for both eyes (n 5 64 patients). Visionimpairment is defined as visual acuity\20/200.bRestricted to those with less than excellent vision at last follow-up(n 5 71 patients).

378 Khitri et al Volume 16 Number 4 / August 2012

and median for factors measured in a continuous scale, and by

proportions for the categorical factors.

For the statistical analysis of visual outcomes at diagnosis and at

last follow-up, visual acuities were categorized into 5 groups: (1)

excellent (20/20 to 20/70, central-steady-maintained); (2) good

(\20/70 to 20/200, central-unsteady-maintained); (3) fair (\20/

200 to 20/800 central-unsteady-unmaintained, central-steady-

unmaintained); (4) poor (\20/800 to light perception, uncentral-

steady-unmaintained, uncentral-unsteady-unmaintained), and

(5) no light perception. The percentage of eyes in each category

was calculated and compared among the different types of

glaucoma.

To examine the risk factors for vision impairment, we de-

fined vision impairment to be a final visual acuity \20/200.

We used univariate and multivariate analysis of risk factors, in-

cluding sex, age at diagnosis, age at surgery, laterality of glau-

coma, type of glaucoma, type of glaucoma surgery, visual acuity

at diagnosis, number of surgeries, and years of follow-up after

diagnosis. The factors were first analyzed by the use of univar-

iate analysis, followed by the multivariate analysis including the

risk factors with P \ 0.10 from univariate analysis. The multi-

variate analyses were simplified by stepwise selection to keep

the statistically significant risk factors with 2-sided P \ 0.05.

The associations of risk factors with vision impairment were

assessed by odds ratio with 95% confidence intervals from

the logistic regression models, with the inter-eye correlation

accounted for by using the generalized estimating equations.5

All the statistical analyses were performed in SAS v9.2 (SAS

Institute Inc, Cary, NC).

Results

A total of 133 eyes of 88 children (50 females [56.8%])with glaucoma were eligible and included; 24 patientswere excluded for insufficient data resultant from signif-icant treatment periods at other hospitals, 3 for age.14 years at diagnosis and 2 for history of trauma.The ocular and systemic presenting characteristics ofthe patients at diagnosis are given in Table 1. Forty-five (51.1%) patients had bilateral glaucoma. The me-dian age at diagnosis was 0.36 years (range, 0-14 years).The median age for each glaucoma group was as fol-lows: 2.53 years (range, 0.07-12.2 years) for aphakic,0.30 years (range, 0.00-2.89 years) for primary congen-ital glaucoma, 0.02 years (range, 0.00-5.93 years) forSturge-Weber glaucoma, 0.29 years (range, 0.00-13.8years) for glaucoma associated with anterior segmentdysgenesis, and 5.50 years (range, 0.14-10.5 years) forothers. The most commonly encountered types of glau-coma were primary congenital glaucoma (36.8%) andaphakic glaucoma (28.6%). Many of the patients pre-sented with other ocular comorbidities, both relatedand unrelated to their diagnosis of glaucoma. Themost common ocular comorbidity was corneal opacifica-tion (39.9%) followed by aphakia postcataract extraction(28.6%). Notably, the vast majority of the eyes (81.2%)had excellent or good vision at time of diagnosis, with

only 2 eyes (1.5%) presenting with visual acuity of nolight perception.

In regard to the therapeutic management of these pa-tients, only 9 (6.77%) were managed medically on anti-glaucomatous medications without a need for surgery(Table 2). Of those eyes requiring surgery, 78 of 124(62.9%) required more than one surgery to control theirIOP. Seven eyes (5.3%) required more than 6 surgeriesduring the course of this study to control their glaucoma;each of these eyes had undergone multiple sessions of en-doscopic cyclophotocoagulation. The most common ini-tial surgery performed was an angle surgery (89 eyes[66.9%]) followed by filtering surgery in 20 eyes(15.0%). When angle surgery was performed, it was al-ways the initial procedure for that particular eye or hadbeen preceded by a previous angle surgery. Patientswho had filtering surgery, particularly a glaucoma tubeshunt procedure, often underwent bleb revision shouldencapsulation have been a cause for failure. Also listedin Table 2 is the last procedure performed for the glau-comatous eyes during the time period studied. Rarely, pa-tients required anterior vitrectomies (isolated from othersimultaneous glaucoma procedures), anterior chamberwashouts for hyphemas, revisions of tube shunts for hy-potony, and synechiolysis—all included in the “other”surgical category in Table 2.

The median follow-up period was 4.95 years (range, 0.3-28 years), with only 6 eyes (4.5%) having less than 1 year offollow-up. The final visual outcomes for the eyes are pre-sented in Table 3. As seen, 80 eyes (60.1%) with glaucoma

Journal of AAPOS

Table 4. Multivariate analysis of risk factors associated with vision impairment at last follow-up

Risk factors No. eyesNo. eyes with visionimpairment, n (%)a Odds ratio (95% CI) P value

Laterality of glaucoma 0.002Bilateral 90 26 (28.9) 1.00Unilateral 43 27 (62.8) 5.46 (1.83-14.0)

Vision at diagnosis 0.02Good ($20/200) 108 72 (33.3) 1.00Poor (\20/200) 25 15 (68.0) 4.68 (1.28-17.1)

Glaucoma diagnosis 0.01Aphakic glaucoma 38 19 (50.0) 1.00Primary congenital glaucoma 49 11 (22.5) 0.25 (0.07-0.87) 0.03Sturge-Weber glaucoma 16 5 (31.3) 0.07 (0.01-0.40) 0.003Glaucoma associated with anteriorsegment dysgenesis

16 10 (62.5) 1.60 (0.34-7.60) 0.56

Aniridia associated glaucoma, uveiticglaucoma, other

14 8 (57.1) 0.64 (0.14-2.84) 0.19

No. surgeries \0.0001\2 55 12 (21.8) 1.002 37 12 (32.4) 3.30 (0.98-11.1) 0.053.2 41 29 (70.7) 32.3 (7.82-133) \0.0001

aVision impairment is defined as visual acuity\20/200.

Table 5. Agreement of visual acuity at diagnosis with visual acuityat final follow-up

Visual acuity at diagnosis

Visual acuity at final follow-up

NLP Poor Fair Good Excellent

NLP 1 1 0 0 0Poor 2 9 2 2 5Fair 0 1 1 0 1Good 3 20 6 11 33Excellent 0 4 3 5 23Percent of agreement 5 45/133 5 33.8%Weighted kappa (95% CI) 5 0.23 (0.16-0.35)

NLP, no light perception.

Volume 16 Number 4 / August 2012 Khitri et al 379

had at least 20/200 vision (excellent and good vision astabulated), whereas 53 (39.9%) had visual acuity \20/200, 6 eyes of which had visual acuity of no light perception.In looking at the patients’ functional visual outcomeor theirvision using both eyes, there were 64 patients with reliablyrecorded visual acuities for both eyes, including all 45 ofthe bilateral glaucoma patients. In these patients withavailable bilateral visual acuity data, 54 (84%) had at leastone eye with good or excellent vision. The primary etiologyfor the patients’ vision impairment (for those withoutexcellent final visual acuity) also was evaluated, and ambly-opia was the most common contributing factor to visionloss (54.9%) followed by glaucomatous optic atrophy(25.4%).Factors associated with vision impairment from multi-

variate analysis are shown in Table 4. Eyes with unilateralglaucoma were 5 times more likely to have vision impair-ment than eyes where the glaucoma was bilateral. Poorvision at diagnosis was significantly associated with an in-creased risk of vision impairment at final follow-up (P 50.02). As shown in Table 5, of the 133 eyes, 45 (34%) had

Journal of AAPOS

visual acuity at final follow-up that was in the same visualacuity category at presentation. Further, final visual acuitiesin 92 eyes (70%) were the same or differed by only one cat-egory from the presenting visual acuity category.

The type of glaucoma also was shown to significantly in-fluence the final visual outcome (Table 4). Eyes with pri-mary congenital glaucoma and Sturge-Weber glaucomawere less likely to have vision impairment than those withother glaucoma diagnoses (P 5 0.01). Although newborn,infantile, and late-recognized primary congenital glau-coma may confer different disease courses, age of diseasediagnosis did not appear to affect final visual prognosis inthe small subset of primary congenital glaucoma patientsstudied (P 5 0.47): visual acuities of $20/200 at final visitwere achieved in 75% patients diagnosed before 6 monthsof age (n 5 35 patients) and in 86% patients diagnosed atage $6 months (n 5 14 patients). Finally, eyes that re-quired more surgeries were more likely to be associatedwith poor visual outcomes. Sex, age at diagnosis and ageat surgery, type of surgery, and years of follow-up wereall not significantly associated with vision impairment(P . 0.05).

The distributions of visual acuity at final follow-up bytype of glaucoma diagnosis are presented in Figure 1.Nearly twice as many eyes with primary congenital glau-coma (34 of 49 [69.4%]) achieved excellent vision at finalfollow-up than the other glaucoma diagnoses (15 of 38with aphakic glaucoma [39.5%]; 6 of 16 with Sturge-Weber glaucoma [37.5%]; 5 of 16 with glaucoma associ-ated with anterior segment dysgenesis [31.3%]; P 50.02). Despite the differences in final visual acuity distribu-tion among the various glaucoma subtypes, the most com-mon etiology for vision impairment at final follow-upremained the same: amblyopia (Figure 2). Among eyesthat achieved a final visual acuity less than excellent,

FIG 1. Visual acuity at final follow-up grouped by pediatric glaucomasubtype (P 5 0.02): excellent (20/20 to 20/70, central-steady-maintained); good (\20/70 to 20/200, central-unsteady-maintained);fair (\20/200 to 20/800, central-unsteady-maintained, central-steady-unmaintained); poor (\20/800 to light perception, uncentral-unsteady-unmaintained, uncentral-steady-unmaintained); and no lightperception (NLP).

FIG 2. Primary etiology for vision impairment in patients with worsethan excellent (\20/70) vision at final follow-up according to glau-coma subtype (P 5 0.20).

380 Khitri et al Volume 16 Number 4 / August 2012

amblyopia accounted for vision loss in 60% of those withprimary congenital glaucoma, 69.6% of those with aphakicglaucoma, 60%of those with Sturge-Weber glaucoma, and54.6% of those with glaucoma associated with anteriorsegment dysgenesis.

Discussion

In our pediatric patients with glaucoma, we found that atleast 60% of the glaucomatous eyes achieved good vision(ie, $20/200) at final follow-up, and nearly one-half(46.6%) of the eyes achieved a visual acuity of $20/70,vision sufficient to qualify for a motor vehicle driving li-cense in most states. Considering the patient’s visionwith both eyes open (functional visual outcome), morethan 84% of patients who could be tested in each eye hadat least one eye with final visual acuity of $20/200. Theseresults suggest outcomes better than those reported by ear-lier case series, which have reported good visual outcomesin 29% to 41% of patients.6-9 The slightly greaterincidence of excellent vision in this study suggests thatprogress that has been made during the last several yearsin the medical and surgical management of pediatricglaucoma, as others have suggested.3,10

We decided to group the various glaucoma subtypes ina slightly different schematic than that reported by Yeungand Walton4 to better reflect the specific subtypes mostcommonly seen at our institution. Patients with primarycongenital glaucoma were all grouped together for thepurposes of this study despite the potential variability ingenotype because genetic testing was not routinely avail-able. It became apparent that primary congenital glau-coma and Sturge-Weber glaucoma conferred a better

visual prognosis than the other subtypes studied. Specifi-cally, only 22.5% and 31.3% of these groups, respectively,had eyes with final visual acuities \20/200. Eyes withother causes of glaucoma with additional concomitant oc-ular abnormalities (aphakia, anterior segment dysgenesis,uveitis, aniridia, and other) have other additional factorscontributing to their poor visual outcome besidesglaucoma.

Poor vision at diagnosis appeared to portend a poorvisual prognosis. This correlation underscores the impor-tance of early diagnosis to improved visual outcomes—inour study, patients who had already sustained considerablevision loss by the time of presentation likely already suf-fered from amblyopia and/or glaucomatous optic neurop-athy. Thus, the goal of effective treatment of pediatricglaucoma may not be to improve patients’ vision but topreserve the vision they present with at diagnosis.

IOP control in pediatric glaucoma is possible withcurrent surgical techniques and medications. The BritishInfantile and Childhood Glaucoma (BIG) Eye study dem-onstrated that IOP control in pediatric glaucoma is suc-cessful in 94% of patients.11 However, good IOP controldoes not necessarily predict a good visual outcome. Ambly-opia proved to be the single most important factor contrib-uting to poor visual outcome in our study. This finding issimilar to visual acuity outcomes associated with other pe-diatric ocular conditions, such as congenital cataracts.12-14

Successful management of pediatric glaucoma cannotfocus exclusively on IOP control but must also ensureadequate amblyopia therapy.

Patients with unilateral glaucoma fared worse thanthose with bilateral glaucoma, just as in their counter-parts with congenital cataracts. In multivariate analysis,

Journal of AAPOS

Volume 16 Number 4 / August 2012 Khitri et al 381

Sturge-Weber glaucoma and primary congenital glaucomawere associated with a lower risk of vision impairment thanother types of glaucoma. However, nearly twice as manyeyes with primary congenital glaucoma achieved excellentvision compared with those with Sturge-Weber glaucoma.This apparently contradictory result is most likely becauseSturge-Weber glaucoma is much more frequently unilat-eral than is primary congenital glaucoma, and unilateralityitself confers a 5-fold greater risk for vision loss, presum-ably related to amblyopia.Patients who required multiple surgeries fared worse

than those who required fewer surgeries. Final visual acuitymay be related to the number of surgeries by the severity ofthe disease (more severe disease causing more vision lossand requiring more surgery) or through amblyopia becauseeach surgery is associated with a certain amount of postop-erative recovery time, which may represent time lost topatching therapy.This study is limited by the difficulty of measuring visual

acuity in the subject population. Visual acuity measure-ments in nonverbal children are less precise than recogni-tion acuity measurements, and over- or underestimates ofacuity are possible.We used a standardized method to con-vert fixation acuity observations to recognition acuity test-ing equivalents. This potential imprecision would mostlyaffect the initial acuity data because 75.9% of final acuitydata was measured by recognition acuity. This study isalso limited by the variability in treatment courses, patientcompliance, and follow-up inherent to any large retrospec-tive study. This study highlights the importance of not onlythe early recognition and aggressive control of elevatedIOP but also the recognition and treatment of amblyopia,

Journal of AAPOS

which is a common cause of vision loss in patients withchildhood glaucoma.

References

1. Kipp MA. Childhood glaucoma. Pediatr Clin North Am 2003;50:89-104.

2. Aponte EP, Diehl N, Mohney BG. Incidence and clinical characteris-tics of childhood glaucoma. Arch Ophthalmol 2010;128:478-82.

3. Biglan AW.Glaucoma in children: Are wemaking progress. J AAPOS2006;10:7-21.

4. Yeung HH, Walton DS. Clinical classification of childhood glau-comas. Arch Ophthalmol 2010;128:680-84.

5. Liang KY, Zeger SL. Regression analysis for correlated data. AnnuRev Pub Health 1993;14:43-68.

6. Kargi SH, Koc F, Biglan AW, Davis JS. Visual acuity in children withglaucoma. Ophthalmology 2006;113:229-38.

7. Haas J. Principles and problems of therapy in congenital glaucoma.Invest Ophthalmol 1968;7:140-46.

8. Haas JS. End results of treatment. Trans Am Acad OphthalmolOtolaryngol 1955;59:333-41.

9. Robin AL, Quigley HA, Pollack IP, et al. An analysis of visual acuity,visualfields, anddisk cupping inchildhoodglaucoma.AmJOphthalmol1979;88:847-58.

10. Bussi�eres JF, Therrien R, Hamel P, Barret P, Prot-Labarthe S. Retro-spective cohort study of 163 pediatric glaucoma patients. Can JOphthalmol 2009;44:323-7.

11. Papadopoulos M, Cable N, Rahi J, Khaw PT, the BIG Eye StudyInvestigators. The British Infantile and Childhood Glaucoma (BIG)eye study. Invest Ophthalmol Vis Sci 2007;48:4100-106.

12. Birch E, Stager DR, Wright W. Grating acuity development afterearly surgery for congenital unilateral cataract. Arch Ophthalmol1986;104:1783-7.

13. Thompson DA, Moller H, Russell-Eggitt I, et al. Visual acuity inunilateral cataract. Br J Ophthalmol 1996;80:794-8.

14. Lewis TL, Maurer D, Brent HP. Development of grating acuity inchildren treated for unilateral or bilateral congenital cataracts. InvestOphthalmol Vis Sci 1995;36:2080-95.